System bugs are caused for three main reasons:
1. Bad Specs
2. Faulty implementatation
3. Incomplete tests and hasty correction of faults.
#1 is by far the worst because it affects #2 and #3 as well and may not be detected during the whole development process.What‘s worse is that in complex developments, this process is cascaded from one system to all of its subsystems.
So here we have an airplane design spec that was most likely not taking into account the aerodynamic behaviour of relocating the engines. I bet the problem was only detected after some test flying of the new airplane. So time runs out, a redesign of the aircraft is now impossible and so a software patch is called for to get better stall protection. The aircraft is not FBW and thus the triple redundancy and voting concept is neither required nor followed. And as all software engineers know, patches sometimes backfire because they have not been designed with the same care as the original system. Ask Bill Gates.
But the patched system is not self-contained, not isolated, it affects and is affected by a lot of other systems and components. If some of these sensors and systems feed wrong air-data, the result is unpredictable probably even for the inventors, but certainly for the poor pilots affected and for the majority of ppruners including me.
Now some of you say „hey that‘s just a little speed bump“, remain cool and switch some breakers and fly away serenely. Thats okay for the Chuck Yeagers and Neil Armstrongs. Those not having been trained as test pilots remain confused and try to fight a plane that does not behave as advertised and tries to kill you and your passengers. They might remember some FAA AD about trim anomalies when they are waiting to pass the pearly gates.
Conclusion: This patch is much worse than the original stability problem. Many aircraft pitch up to a stall if uncorrected after full power is applied, at least mine does. That‘s no issue. It baffles me that a device that can trim a jet fully nose-down without pilot input - caused by some erronous airdata - could ever be conceived and then certified and the facts about this box of Pandora not being passed on to the type-rated pilots and training facilities.